How do we quantify life span brain health in individuals and in populations?

How do we integrate modalities and organ systems?

need unbrain example here

ANTsR \(\rightarrow\) new insight via quantification

Good software should fade into the background … however …

As is common in science, the first big breakthrough in our understanding … [came from] an improvement in measurement.

> Daniel Kahnemann, Thinking, Fast and Slow (2011)

SyN video

Theory + evaluation + reproducibility

Tools you can use for imaging science

• Core developers: B. Avants, N. Tustison, H. J. Johnson, J. T. Duda

• Many contributors, including users …

• Multi-platform, multi-threaded C++ stnava.github.io/ANTs

• Developed in conjunction with http://www.itk.org/

• R wrapping and extension stnava.github.io/ANTsR

• rapid development, regular testing \(+\) many eyes \(\rightarrow\) bugs are shallow

A long history of research

Rigorous transformation definition is key

ANTs and ITK are developed together: see B. B. Avants et al. (2014), Tustison and Avants (2013), Tustison et al. (2010) and more …

Key definitions

• physical space

• transformation definition aware of physical space

• optimization space consistent with above

• unit testing

bart

open source

built on ITK—probably the most well-vetted medical image analysis package in the world B. B. Avants and Tustison (2014)

competitions

• Klein 2009: Brain Registration (ANTs)
• Murphy 2010: Lung Registration (ANTs)
• SATA 2012: Multi-Atlas Segmentation (ANTs+JointLabelFusion)
• SATA 2013: Multi-Atlas Segmentation (ANTs+JointLabelFusion)
• BRATS 2013: Multivariate Brain Segmentation (ANTsR)
• Yushkevich’s Hipp Atlas: ( hippocampusubfield.com )
• TBA: BOLD decoding (ANTsR)
• Substantial work with DTI ( Camino developer in house )
• STACOM2014 ?

papers

• registration : ANTs vs. everything else Klein et al. (2009)

• segmentation : Atropos vs. SPM, etc.

• bias correction : N4 vs N3

• cortical thickness : ANTs vs. FreeSurfer Tustison et al. (2014)

• compatibility with R

Registration

N4

• N3 (developed at the Montreal Neurological Institute) has been the gold standard for bias correction—used in important projects such as ADNI

• N3 is a set of perl scripts that works natively with the MINC file format which we tried to incorporate into an ANTs processing pipeline.

• We had so much trouble converting back and forth between ITK-compatible Nifti format and MINC that Brian suggested we try to implement N3 in ITK.

• I had some experience with B-splines and added some other tweaks giving birth to N4.

Atropos

• Similar to our experience with N3, we tried to incorporate FAST (from the FMRIB at Oxford) into an ANTs processing pipeline.

• Phil (?) pointed out the difficulty of incorporating priors into FAST.

• Related, Brian went to a segmentation-related worksop at MICCAI and aired publicly his disappointment that so much of what had been developed in the community over the last 20+ years has not been made publicly available. “What’s wrong with you people!”

• 3-tissue algorithm in ImageMath —> multivariate, n-class Atropos

KellySlater –> KellyKapowski

Software engineering

ants gource

Problems

There are several problems and shortcomings to this analysis.

• Customizable for specific problems but not too specific
• Latest theoretical advances in registration not yet wrapped for users

Strengths

Some strengths include relatively few assumptions, a flexible implementation and open-science approach.

References